CSTR - Coinstar

[Picasso]

were having this discussion again??

 

Dividend is far inferior to buybacks in this situation. If you think buybacks are bad, you think that the stock is expensive and you should not own it. If you are a valueinvestor and you think you own cheap stocks, and you want a dividend, your basicly saying your stocks are expensive.

 

Also read the thread, and read the CSTR write ups on VIC they explain the business well. It iwll never go to zero. probably 100m$ is the lowest it can go. coinstar will place a 80m$ FCF bottom in this at year 5. Unless we stop using cash at year 5. That seems highly unlikely.

 

If they put a dividend sometime in the next 2 years if the stock is still low, stock would probably rerate to 90$ (9x multiple on 200m$). Sure you make a decent return.

 

They buy back another 350m$ at about 60$ a share, they will have about 14m shares outstanding. Stock rerates to a 9x multiple , you get 128$ per share. Even if you reinvest the dividend, you will not get that return with dividends. Your return would still be 108$.

 

They buy back 700m$ of shares at 70$? 10m shares outstanding. Let's say they generate 200m$ in year 3 and a 9x multiple = 180$ per share. Would have been 130-140$ a share with dividend.

 

Even all your dividends reinvested at 70$ you will get a lower return then with buybacks.

 

Agreed you will get a lower return with dividends but you can deploy it into other more durable areas.  There is a higher degree of certainty of getting the cash out of the business.  The buybacks hold less value because of risks in the business.  If those risks turn out to be nonexistent then you are getting a steal here because they can buyback shares at 4x FCF.  But if they turn out to be real risks you are likely to end up with a very poor return.  At least based on my worst case scenario.

 

Now if you are right and OUTR can always do good cash flows it will be worth something considerable at the end even with $100 million of FCF.  But this is a leveraged stock, not a debt-free entity.  Just because someone on VIC thinks they will steady state with all the debt does not mean it will be the case. 

 

By the way, I am not bearish on this stock.  I am simply trying to figure out the margin of safety and low end of the expected returns.  As I mentioned in an above post, this could be a 10x easy.  But it can almost just as easily be a disaster.

[KCLarkin]

Anyways, Charlie Munger once said that to find great stocks you have to "look at the cannibals," companies that eat themselves over time by buying back their own shares.  If you think about OUTR in that manner, as a cannibal, then the calculus for owning this stock starts to change (at least it did for me). 

 

Be careful. OUTR does not pass the second filter of the Munger/Buffett four filter test. Are we reasonably sure the business will be worth more in 10 years than it is now? It might be an attractive purchase at these prices, but this is not a Modern Munger stock. Do not use Munger to justify a purchase.

 

I suggest plugging these numbers into excel, with the assumption that Coinstar free cash flow will remain relatively constant over the next ten years (increase consistent with inflation), the assumption that DVD rentals will wither away and thus Redbox free cash flow will decline by 10%, every year, over the next ten years, and the assumption that ecoATM, SampleIT, and Coinstar Exchange are absolutely worthless.

 

Have you built this model? There is a lot of theoretical discussion in this thread but I haven't seen any evidence that people are actually putting enough effort into this idea.

 

For those who prefer dividends to buybacks, you are implicitly saying the stock is currently overpriced. I can't imagine any math that would make purchasing undervalued shares result in destruction of value. I think HLF is overpriced. I don't care whether they do buybacks or dividends.

 

Buyback versus dividend argument is a waste of time. Is this overpriced or underpriced?

[Picasso]

See I think people are mistaking buybacks in nondurable businesses as a difference between a cheap or expensive stock.  It does not matter because no one knows the durability of HLF, OUTR, and some even think IBM is not durable.  OUTR can be a cheap stock if things go okay, it can be an expensive stock if the business deteriorates.

 

I would just like to know what my returns will be if the business deteriorates.  It will turn out the stock is expensive but it did not mean it was expensive when we started out looking at shares today.  Whether shares or cheap or expensive years from now based on purchasing shares today is to be seen.

 

Was SHLD expensive at $100?  In hindsight it was but not at the time.  The deterioration in the business made it expensive and that is a real risk to companies like this which eat up their own shares.  That said I know I am probably not going to get anywhere with the buyback discussion so I will stop flogging that dead horse.

[shhughes1116]

Have you built this model?

 

Yes.

 

 

[shhughes1116]

Although to clarify, I don't think a model is necessary for this equity. 

 

The short thesis seems to rely on predictions that: (1) RedBox will collapse over the next five years; and (2) Coinstar will collapse over the next 5 years; and (3) Coinstar Exchange will be worthless; and (4) ecoATM will be worthless; and (5) SampleIT will be worthless.  Think about those predictions from the perspective of probability.  The current stock price seems to reflect the idea that all five of those predicted "outcomes" are going to happen.  What is the actual probability that all five of those predicted "outcomes" will occur in the next five years?  The probability that one of predicted "outcomes" is incorrect far exceeds the probability that all five predicted "outcomes" are correct.  And if one of the predicted "outcomes" is incorrect, then the current price of the equity is cheap relative to the discounted value of the future cash flows. 

[Picasso]

Although to clarify, I don't think a model is necessary for this equity. 

 

The short thesis seems to rely on predictions that: (1) RedBox will collapse over the next five years; and (2) Coinstar will collapse over the next 5 years; and (3) Coinstar Exchange will be worthless; and (4) ecoATM will be worthless; and (5) SampleIT will be worthless.  Think about those predictions from the perspective of probability.  The current stock price seems to reflect the idea that all five of those predicted "outcomes" are going to happen.  What is the actual probability that all five of those predicted "outcomes" will occur in the next five years?  The probability that one of predicted "outcomes" is incorrect far exceeds the probability that all five predicted "outcomes" are correct.  And if one of the predicted "outcomes" is incorrect, then the current price of the equity is cheap relative to the discounted value of the future cash flows.

 

But 2/3rds of cash comes from Redbox and the other 1/3rd from Coinstar. Their other ventures so far are losing money.  You really only need a hit to either business to throw a monkey wrench in the whole operation.

[KCLarkin]

See I think people are mistaking buybacks in nondurable businesses as a difference between a cheap or expensive stock.  It does not matter because no one knows the durability of HLF, OUTR, and some even think IBM is not durable.  OUTR can be a cheap stock if things go okay, it can be an expensive stock if the business deteriorates.

 

I would just like to know what my returns will be if the business deteriorates.  It will turn out the stock is expensive but it did not mean it was expensive when we started out looking at shares today.  Whether shares or cheap or expensive years from now based on purchasing shares today is to be seen.

 

Was SHLD expensive at $100?  In hindsight it was but not at the time.  The deterioration in the business made it expensive and that is a real risk to companies like this which eat up their own shares.  That said I know I am probably not going to get anywhere with the buyback discussion so I will stop flogging that dead horse.

 

The uncertainty and durability should be factored into the valuation. Again, the problem isn't the buybacks it is the uncertainty and the durability, especially given the debt. Even if they are paying out all the FCF as dividends, you need to make a bet on how long that FCF lasts before the debt holders take over the company.

[yadayada]

Except OUTR is durable. If it isn't you should not bet on it. Redbox will probably exist in some form 10 years from now, and Coinstar will generate more cash 10 years from now. That is basicly a monopoly with a rock solid moat as long as people use cash. The assumption that they will generate 0 FCF in 5 years is way too pessimistic and extremely unlikely.

 

buybacks

So coinstar generating like 90m$ 5 years from now (currently about 70m$) looks likely with the new machines being installed and tracking inflation. And judging by it's past 20 years. And redbox FCF declines 15% a year starting next year. So that is about 100m$ for this year. 187m$ +160+135+114+110. And I think there will always be some niche, so it probably stays even at some point. So this is about 800m$. And let's give that remaining 110$ a multiple of 4x. Note that redbox can lose 50% of revenue before losing money. So it is really easy to scale this down. No high fixed costs like blockbuster.

 

So Let's say 500m$ went to buy backs and 300m$ to new ventures. And coinstar paid off 300m$ of debt with the ~400m$ they generated. And also spent 100m$ on new ventures, so 400m$ wasted on new ventures.

 

So 600m$ of debt, FCF is now 190m$ with coinstar going up and Redbox going down. NPV of coinstar is about 12x 90m$ = 1080 + 300m$ of redbox (lets be very conservative here). So a 7.2x fair value multiple.

 

No new ventures, they failed.

 

So that would be ~7.5 million shares at an average of 65$. or about 12.5m shares outstanding. 1380/12.5 = 110$ per share.

 

 

Dividends

Now the exact same scenario except a dividend (and you did not reinvest because you thought the business was dying, because otherwhise preferring dividends in this case is pretty irrational), Instead you reinvested the dividend at 15% somewhere else. Since they did not pay out dividend at once, let's say on average 4 years instead of 5. or about 43$ a share.

 

1380/20m shares (same thing as above, 600m$ in debt etc), = 69$ + 43$ = 112$ per share. Or 94$ per share if you did not reinvest.

 

So in bear case you are basicly breaking even with dividend (assuming you can actually invest somewhere else and get 15%) and buyback. I don't think coinstar goes away. Cash does not go away, it has it's use. And as long as people use cash, coinstar will be around making money with a rock solid monopoly.

 

And I think taking scenario where redbox starts losing 15% each year and then there is only 300m$ of residual value is also pessimistic. And ofcourse assuming new ventures are a disaster and 400m$ is wasted on them is also pessimistic.

 

If your more optimistic in a base case then buybacks will be clearly superior. Unless you can reinvest dividends at a really high rate somewhere else.

 

So that is kind of the point here. Even in a pretty bearish scenario in 5 years, the market cap should be higher then 1 billion$. So that is why buybacks are good here. Redbox falling off a cliff within 1-2 years is highly unlikely for the reasons mentioned here. If the market cap in a few years (so the NPV of cash flows from then on) is higher then the current market cap, buybacks are almost always going to be a better idea. Especially if that scenario makes pessimistic assumptions.

 

The only real risk here is that they waste like 4-600m$ on new ventures that turn out to be a bomb. And that not enough debt is being paid off.

 

And finally your not holding this stock 5 years. Within 2 years the market will probably be rational here and it will rerate to some appropriate multiple and it won't be worth holding on to any longer.

 

If I had to guess, 90% of the time, NPV of cash flows will probably be somewhere between a billion and 5 billion. Most likely scenario somewhere around 2-3 billion (kind looks like a normal distribution). So I want them to buy back as much shares as possible at these prices.

 

This is really a matter of realistically looking how much cash is left in Redbox, and how strong the coinstar business is. And there are not really any signs that redbox will start collapsing, except for some superficial "omg dvd is dead" despite a lot of good reasons why that is not true.

[Picasso]

Is there any reason Liberty Media does not take a big stake in this business and milk that free cash flow and likely better manage the company? I mean they invested in Barnes and Noble for goodness sakes. This seems like such low hanging fruit for them aside from the $2B EV, but they can approach it like a public LBO.

 

Does anyone have extra insight into what happened with JANA's investment? There must be a reason why a sale did not take place and they sold all their shares since it is still cheap on a public equity basis.

[KCLarkin]

Although to clarify, I don't think a model is necessary for this equity. 

 

The short thesis seems to rely on predictions that: (1) RedBox will collapse over the next five years; and (2) Coinstar will collapse over the next 5 years; and (3) Coinstar Exchange will be worthless; and (4) ecoATM will be worthless; and (5) SampleIT will be worthless.  Think about those predictions from the perspective of probability.  The current stock price seems to reflect the idea that all five of those predicted "outcomes" are going to happen.  What is the actual probability that all five of those predicted "outcomes" will occur in the next five years?  The probability that one of predicted "outcomes" is incorrect far exceeds the probability that all five predicted "outcomes" are correct.  And if one of the predicted "outcomes" is incorrect, then the current price of the equity is cheap relative to the discounted value of the future cash flows.

 

In your worst case scenario above, I think the stock is worth about $5 per share today:

 

- 15% discount rate

- Assume company retains all FCF till end

- $200 M FCF x 5 years

- $1000 M terminal debt payment

- $0 terminal value

 

So I don't think the short thesis is that drastic. I think the short thesis is more like:

 

(1) RedBox will collapse over the next 10 years;

(2) Coinstar will collapse over the next 10 years;

(3) New ventures will have no value

 

[shhughes1116]

I used 5 years because someone suggested that time frame in an earlier post.  And I don't disagree with the broad math you laid out in your post.

 

My intention was to point out that the short thesis relies on everything going wrong for Outerwall.  However, the probability of collapse/failure for each of the five predictions identified in my previous post, as outlined by the union of probabilities for all five events -->  [P(1) u (2) u (3) u (4) u (5)] - is FAR less than the probability that something goes right for Outerwall.       

[Picasso]

I used 5 years because someone suggested that time frame in an earlier post.  And I don't disagree with the broad math you laid out in your post.

 

My intention was to point out that the short thesis relies on everything going wrong for Outerwall.  However, the probability of collapse/failure for each of the five predictions identified in my previous post, as outlined by the union of probabilities for all five events -->  [P(1) u (2) u (3) u (4) u (5)] - is FAR less than the probability that something goes right for Outerwall.     

 

I wasn't saying cash flow will go to zero in five years.  Sorry if it came off in that manner.  I think it would be a gradual process which may start at year five and end year 8 with minimal or zero cash flows.  We can play with the years but if you are assigning a decent discount rate to that scenario, it will not matter much whether it implodes in year 8 or 10.

 

Probability wise, I think it is high that you see a deterioration in the cash flow some years out.  The degree of it is debatable but I am simply trying to figure out the floor on the stock if the wheels fall off the bus.

[yadayada]

You have to analyze coinstar and redbox then. Because coinstar is extremly likely to be around 10 years from now, and veyr likely bigger then it is now. It is a completely different animal then Redbox with a rock solid monopoly. It is almost impossible to break in that business since there is only room for one, and you need scale to compete.

 

Cash is not really going away. The % of unbanked people is higher now then a few years ago, and cash will always have it's uses. Especially not going away with the likely eroding middle class.

 

And there is also some operating leverage in coinstar. Plus looking at 2008-09, it is quite recession proof, and it will likely outpace inflation.

 

[hildenbrand]

Great discussion,

 

I think Picasso brings up an interesting and valid point regarding dividends.  Seems there’s a near consensus (right or wrong) that EVENTUALLY redbox goes away.  Studios are constantly experimenting (Snowpiercer was recently released on VOD the same day as theatrical) and redbox is potentially affected by studios actions - i.e. IF studios priced VOD at $2 redbox would be in a world of hurt.  Doesn’t seem crazy to think at some point redbox is sideswiped by studios experimenting and changing consumer tastes - there are a lot of moving parts.

 

As of today redbox looks fairly solid and very profitable, and no doubt cheap if there’s a long stable runway.  However it’s always hard to predict future changes in an industry like this with many moving parts in the value chain - broadband speed/costs, studio windowing, VOD pricing and streaming services.  In my opinion handicapping all the key issues is difficult looking 10 years out, though it’s hard to see things changing in the immediate future.  Because of this dynamic I’d prefer a policy of dividends instead of buybacks as a measure of safety in case sudden change impairs value, thus making buybacks look foolish in hindsight. 

 

In other words the FCF yield today is very attractive, yet at some point a large portion of the cash flow likely disappears, whether that’s in 5 yrs or 15 yrs is very hard for me to say with conviction, therefore paying out 85% in dividends seems the most prudent even if it proves sub-optimal.

 

 

[krazeenyc]

FWIW, management has said their plan is to do buybacks up to a certain point -- then their plan is to shift to dividends.

 

 

No position in OUTR. 

[yadayada]

you cannot say that because of one unlikely scenario where NPV of future cash flows is less then current EV, buybacks are bad.

 

IF studios priced VOD at $2 redbox would be in a world of hurt.  Doesn’t seem crazy to think at some point redbox is sideswiped by studios experimenting and changing consumer tastes - there are a lot of moving parts.

 

Yeah but that does not happen. If they wouldl do that, and get a very optimistic number of rentals for new releases, they will still make much less then they do now. The studio's controlling the release window basicly keeps demand high for higher priced new releases.

 

So in 80% of the outcomes, the stock is cheap now, you should still do buybacks. Doesn't matter that in 20% of the cases buybacks are a bad idea. The added value in the cases where the stock is cheap gives more then enough enhanced returns to make up for the destroyed value in 20% of the scenario's.

 

The moment you think that coinstar will be in danger long term, and that other ventures are a cigar butt, AND the NPV on average of those over the next 10 years is less then the EV, then you should start paying dividends.

 

So NPV over 20 years of coinstar is likely somewhere between 2-2.5 bn$ (given they grow a few % a year). So let's say 1.6 bn$ to be safe. It is likely Redbox is at least 7-800m$, and possibly somewhere up to 2-3 billion$.

 

And then there are new ventures that could add possibly several billion$, or negative value. So if you average that out, then NPV is more then the current 2 billion$ ev in most cases. And if you take out 1 billion$ for coinstar to take out the debt, more then the 1 billion$ equity value over even the next 10 years.

 

So possibly equity value could be 2-4 billion$.. So you have to also consider the mistake of NOT buying back in this scenario. Because buying back truckloads in this scenario really puts your return on steroids vs paying dividends.

 

 

 

Finally in a liquidation, buybacks could still be the correct option. Let's say a company has a liquidation value if 100 million and liquidates its shares in 2 parts of 50 million each after all transaction costs. Market values them at 50 million$ with 10 million shares.

 

Let's say they buyback half the shares at 5$ a share for 25 million. And pay the second half in dividend. Now the ones who held on, that is 5 million shares, that will get 75 million$.

 

Let's say I hold 1000 shares, bought at 5$. I did not sell any during the buyback. now after buy back I get 75 million/ 5 million = 15$ per share. Or 15k$ total.

 

In the dividend without a buyback first, I get 100 million $ / 10 million = 10$ per share. Or 10k$.

 

So even with a cigar butt that is badly priced, you still want them to buy back as long as the stock is cheap. And the reason debt does not matter much is because coinstar will still be around, probably making more money 10-15 years from now.

 

Im done typing up huge walls of text now lol.

[yadayada]

I guess you can summarize that post as:

 

If NPV of future cash flows over likely remaining life of an asset is greater then the fully dilluted market cap + debt, buy backs ALWAYS create more value for shareholders that hang on, then just paying dividends. The moment the market price starts reflecting the value of remaining cash flows, then dividends are better. And the ones who held one are rewarded with a larger return.

 

So that is why you need to have an opinion on coinstar in this case. And that is also why you need to look beyond 10 years for coinstar to have a large margin of safety.

[Picasso]

But a 1/5 chance of a zero dollar terminal value is a bad chance.  That is literally playing Russian roulette with a 5 chamber pistol.  Especially if that outcome involves the return of capital into repurchases.  You will not see that capital again.

 

Also there is no liquidation value in this stock, it is just the cash flow.  It would be like trying to liquidate Phillip Morris.  They just generate a lot of cash but there are no significant net assets.

 

In my opinion there is no real margin of safety in this stock if things get bad.  But there is a lot of upside if things remain stable.

 

But you argue Coinstar has a moat and will represent the value of OUTR in a worst case scenario.  I need to look more into that business to have a better idea of whether that is true or not.

[yadayada]

Yes but the NPV of future cash flows is an average of different scenario's. In the end there will be one outcome, but to price the risk your taking, you have to estimate what different outcomes there are, and how likely it is they will happen. And then take an average. And if the NPV of that average is higher then the market cap (preferably by a wide margin, which i think it is) then buybacks are better.

 

If 20% of the time this thing goes to zero, and 80% of the time it is worth 2 billion$, then they should buy back, your average return will be higher. ALlthough your risk will be higher too, as dividends would mitigate the downside scenario somewhat. But I don't think there is a 20% chance this goes to zero. I think it's much much lower. Also in the downside scenario, that would have to show quickly. They cannot pay out a  lot of dividends anyway in the going to zero scenario. So I doubt your risk will be lowered enough to sacrifice the higher return.

 

And you evaluate this each year. So let's say in year 3 market cap is now 1.5 billion$ , and Redbox is going down rather quickly, you might say, well NPV is now getting closer to market value, let's pay out dividends.

 

Liquidation value is not relevant here. You estimate NPV of future cash flows (so this is an average of all the outcomes) and compared that to the enterprise value before every buy back. 

 

And if there is no margin of safety if things go bad (you need to assume coinstar will go bad, any arguments for this? I havent found good ones so far) that is not relevant. You can lose with every stock, as long as the upside makes up more then enough for the downside then you will on average beat the market. So same arguments goes for the expected value of dividends vs buybacks. Unless there is a larger chance they will lose, and a small chance they win big. In that case I would probably prefer dividends too in the mean time.

 

Also as for studying coinstar, there is a lot of good stuff in the VIC write ups. There are like 5 of them, note to search for CSTR and not OUTR.

[shhughes1116]

Determining whether a company has a durable advantage is a tricky task. 

 

With respect to Coinstar, I'd offer up the following:

1.  Scale of operation ensures that Coinstar is the low-cost operator.  As more kiosks are added, the ratio of fixed costs per kiosk declines.  A company with the scale to compete with Coinstar on cost (i.e. a money center bank or national retailer) probably won't bother given the amount of revenue/income at stake.  And if they choose to compete, it would probably be easier to simply buy Outerwall.  A smaller-sized company for which the revenue/income is meaningful probably won't have the wherewithal to compete on cost with Coinstar given the large network of existing Coinstar kiosks.     

2.  Even banks are getting out of the business of counting coins.  Although I have not seen this with my own eyes, I am told that a number of bank branches have rented space to Coinstar.  They would rather receive the rent money from the kiosk than pay staff to count coins for customers. 

3.  As yadayada said, I don't see hard currency going away any time soon.  Someone, or something, will have to count it... 

4.  Countries in which Coinstar is expanding rely more on coins than paper currency (i.e. UK and Ireland).  In the UK, the smallest denomination for paper is 5 pounds.  Coins are used for 1 pound and 2 pound denominations.  Same with the Euro, which many retailers accept in the UK.     

[KCLarkin]

Also as for studying coinstar, there is a lot of good stuff in the VIC write ups. There are like 5 of them, note to search for CSTR and not OUTR.

 

Thanks! I never thought to search for CSTR.

[Guest wc]

I disagree with for now on cheap online rentals. If you look at dvd retail sales of top titles, they are flattening out .

 

Top 100 DVD titles sales in the US for retail is 2.3 billion$ in 2011, 2.3 billion in 2012 and 2 billion in 2013.

 

But bluray top 100 sales were 1.7 billion$ in 2013 and still growing in 2014 so far for about 10% first half. They have roughly 50% margins on these things, and it isn't really slowing down very fast.

 

And if you look at top 20 titles then dvd sales are roughly flat at 1 billion$ the last 3 years, and blu ray is  at 1 billion$ and growing in 2013.

 

The only dvd sales that are collapsing are older and more unknown titles. So for top 100 titles (most new release titles each year) , that is a 1.8 billion$ cash cow.

 

Mind posting where you're finding these statistics? Been looking for similar numbers with no luck.

[yadayada]

http://www.the-numbers.com/weekly-dvd-sales-chart

 

I put the numbers in excel and added them for a write up i did not post . I think there is one mistake i made in adding them up0 though. They are in attachment.

 

Im hopefull for this year though. Compared to last year, so far we have guardians of the galaxy, transformers and Xmen that were all hits, And decent movies actually. They will probably be on redbox by Q4?

[shhughes1116]

I think the fourth quarter will be alright given what I see in the local RedBox. 

 

I think the third quarter is going to be a lot like the 2nd quarter...below expectations given the weak movie slate. 

[Travis Wiedower]

Q3 looks good--margins up across the board, ecoATM roll out keeping pace, decent amount of repurchases. Won't have time to listen to the earnings call until tomorrow.

 

I revisited the 2015/16 box office release schedules today--should be great. 2014 is probably going to end up being the worst box office year over a ~5 year period.